Hierarchical Task Network (HTN) planning is a powerful paradigm for modeling complex decision-making problems through task decomposition. However, traditional HTN planning assumes deterministic task costs, limiting its applicability in real-world scenarios where execution outcomes are uncertain. Risk-aware HTN planning introduces the use of utility theory to rank plans in domains with probabilistic costs, according to an agents risk attitude. However, the utility of a plan is only evaluated during initial planning. Agents with dynamic risk attitudes lack a mechanism to reevaluate the plan during execution, if unexpected high or low costs are incurred. We address this gap, by developing an approach for replanning in risk-aware HTN planning. We introduce a framework to model risk-aware replanning problems. Additionally, we present an algorithm to solve such problems, allowing agents to modify the plan during execution to represent their change in risk-attitude. We implement this solution in Risk Aware Replanning in JSHOP2 (RARJSHOP2), an extension of the Java Simple Hierarchical Ordered Planner 2 (JSHOP2), supporting probabilistic costs and automatic replanning, based on changes in plan utility due to incurred costs. Evaluation in an electric vehicle routing domain demonstrates that our approach successfully adapts plans according to changing risk preferences, though at increased computational cost.